PEuropean A Gunatillake Cells &and R MaterialsAdhikari Vol. 5. 2003 (pages 1-16) DOI: 10.22203/eCM.v005a01 for tissue ISSN engineering 1473-2262

BIODEGRADABLE SYNTHETIC POLYMERS FOR Pathiraja A.Gunatillake and Raju Adhikari CSIRO Molecular Science, Bag 10, Clayton South MDC, Vic 3169, Australia

Abstract Introduction

This paper reviews biodegradable synthetic polymers fo- Biodegradable synthetic polymers offer a number of ad- cusing on their potential in tissue engineering applica- vantages over other materials for developing scaffolds in tions. The major classes of polymers are briefly discussed tissue engineering. The key advantages include the abil- with regard to synthesis, properties and biodegradability, ity to tailor mechanical properties and degradation ki- and known degradation modes and products are indicated netics to suit various applications. Synthetic polymers based on studies reported in the literature. A vast major- are also attractive because they can be fabricated into ity of biodegradable polymers studied belongs to the poly- various shapes with desired pore morphologic features family, which includes and conducive to tissue in-growth. Furthermore, polymers can polylactides. Some disadvantages of these polymers in tis- be designed with chemical functional groups that can sue engineering applications are their poor induce tissue in-growth. biocompatibility, release of acidic degradation products, Biodegradable synthetic polymers such as poor processability and loss of mechanical properties very poly(), poly() and their copoly- early during degradation. Other degradable polymers such mers, poly(p-dioxanone), and of trimethyl- as , polyanhydrides, polyphosphazenes, and ene carbonate and glycolide have been used in a number are also discussed and their advantages and of clinical applications (Shalaby, 1988; Holland and disadvantages summarised. With advancements in tissue Tighe, 1992; Hayashi , 1994; Kohn and Langer, 1997; engineering it has become necessary to develop polymers Ashammakhi and Rokkanen, 1997). The major applica- that meet more demanding requirements. Recent work has tions include resorbable sutures, systems focused on developing injectable compositions and orthopaedic fixation devices such as pins, rods and based on poly (propylene fumarate) and poly (anhydrides) screws (Behravesh et al., 1999; Middleton and Tipton, to meet these requirements in orthopaedic tissue engineer- 2000). Among the families of synthetic polymers, the ing. Polyurethanes have received recent attention for de- have been attractive for these applications be- velopment of degradable polymers because of their great cause of their ease of degradation by of ester potential in tailoring polymer structure to achieve me- linkage, degradation products being resorbed through the chanical properties and biodegradability to suit a variety metabolic pathways in some cases and the potential to of applications. tailor the structure to alter degradation rates. Polyesters have also been considered for development of tissue en- Key Words: Biodegradable polymers, tissue engineering, gineering applications (Hubbell, 1995; Thomson et al degradation, injectable polymers 1995a, Yazemski et al. 1996; Wong and Mooney, 1997), particularly for bone tissue engineering (Kohn and Langer, 1997; Burg et al., 2000). Attempts to find tissue-engineered solutions to cure orthopaedic injuries/diseases have made necessary the development of new polymers that meet a number of de- manding requirements. These requirements range from the ability of scaffold to provide mechanical support dur- ing tissue growth and gradually degrade to biocompatible products to more demanding requirements such as the ability to incorporate cells, growth factors etc and pro- vide osteoconductive and osteoinductive environments. Furthermore, the development of in-situ polymerizable Address for correspondence: compositions that can function as cell delivery systems Pathiraja A. Gunatillake in the form of an injectable liquid/paste are becoming CSIRO Molecular Science, Bag 10, increasingly attractive in tissue engineering applications. Clayton South MDC, Many of the currently available degradable polymers do Vic 3169, not fulfil all of these requirements and significant chemi- Australia cal changes to their structure may be required if they are Telephone number: 61 3 9545 2501 to be formulated for such applications. E-mail: [email protected] Scaffolds made from synthetic and natural polymers,

1 P A Gunatillake & R Adhikari Polymers for tissue engineering

and ceramics have been investigated extensively for or- Major classes of degradable polymers thopaedic repair. This approach has advantages such as the ability to generate desired pore structures, matching Polyesters size, shape and mechanical properties to suit a variety of A vast majority of biodegradable polymers studied be- applications. However, shaping these scaffolds to fit cavi- long to the family. Table 1 lists the key poly- ties/defects with complicated geometries, bonding to the mers in this family. Among these poly(α-hydroxy acids) bone tissues, and incorporating cells and growth factors, such as poly(glycolic acid) (PGA), poly(lactic acid) (PLA), and requirement of open surgery are a few major disad- and a range of their copolymers have historically com- vantages of this approach. prised the bulk of published material on biodegradable A material that can be used as a scaffold in tissue polyesters and have a long history of use as synthetic bio- engineering must satisfy a number of requirements. These degradable materials (Shalaby, 1988; Holland and Tighe, include biocompatibility, to non toxic 1992; Hayashi , 1994: Kohn and Langer, 1997; products within the time frame required for the applica- Ashammakhi and Rokkanen, 1997) in a number of clini- tion, processability to complicated shapes with appropri- cal applications. These polymers have been used as su- ate porosity, ability to support cell growth and prolifera- tures (Cutright et al., 1971) plates and fixtures for frac- tion, and appropriate mechanical properties, as well as ture fixation devices (Mayer and Hollinger, 1995) and maintaining mechanical strength during most part of the scaffolds for cell transplantation (Thomson et al. 1995b). tissue regeneration process. Development of a degrada- ble polymer composition that can be injected Poly(glycolic acid), poly(lactic acid) and their arthroscopically has number of advantages in tissue en- copolymers gineering as against prefabricated scaffolds. A major ad- Poly(glycolic acid) (PGA) is a rigid thermoplastic mate- vantage would be the possibility of administering the rial with high crystallinity. (46-50%). The glass transi- arthroscopically avoiding surgery in many cases. It also tion and melting temperatures of PGA are 36 and 225ºC, has the advantage of filling cavities with complex respectively. Because of high crystallinity, PGA is not geometries, and to provide good bonding to tissue. Cells, soluble in most organic solvents; the exceptions are highly growth factors and other components to support cell fluorinated organic solvents such as hexafluoro isopro- growth could also be incorporated with the gel. Such poly- panol. mer systems also have the potential to be formulated to Although common processing techniques such as ex- generate porous structure upon curing to facilitate nutri- trusion, injection and compression moulding can be used ent flow to cells during growth and proliferation. Fur- to fabricate PGA into various forms, its high sensitivity ther, such polymers may be useful in pre-fabricating scaf- to hydrolytic degradation requires careful control of folds with complex shapes with appropriate pore struc- processing conditions (Mikos and Temenoff, 2000, Jen tures with biological components incorporated. However, et al. 1999). Porous scaffolds and foams can also be fab- in addition to the main requirements mentioned above, ricated from PGA, but the properties and degradation an injectable polymer composition must meet the follow- characteristics are affected by the type of processing tech- ing requirements to be useful in tissue engineering appli- nique. Solvent casting, particular leaching method and cations. Ideally the prepolymer should be in liquid/paste compression moulding are also used to fabricate PGA from, sterilizable without causing any chemical change, based implants. and have the capacity to incorporate biological matrix The preferred method for preparing high molecular components. Upon injection the prepolymer mixture weight PGA is ring-opening polymerization of glycolide should bond to biological surface and cures to a solid and (Figure 1), the cyclic dimer of glycolic acid (Hollinger et porous structure with appropriate mechanical properties al. 1997, Sawhney and Drumheller, 1998), and both so- to suit the application. The curing should be with mini- lution and melt polymerization methods can be used. The mal heat generation and the chemical reactions involved common catalysts used include organo tin, antimony, or in curing should not damage the cells or adjacent tissues. zinc. If stannous octoate is used, temperature of approxi- The cured polymer while facilitating cell in-growth, pro- mately 175ºC is required for a period of 2 to 6 hours for liferation and migration, should ideally be degraded to polymerization. Although it is possible to synthesize these biocompatible components that are absorbed or released polymers by acid-catalysed polycondensation of respec- from the body. tive acids, the resulting polymers generally have a low The aim of this paper was to briefly review the major molecular weight and often poor mechanical properties classes of biodegradable polymers and their potential in (Agrawal et al. 1997). developing injectable polymer systems in tissue engineer- The attractiveness of PGA as a biodegradable poly- ing. The review was focused on synthesis, potential ap- mer in medical application is that its degradation prod- plications, biocompatibility and biodegradation of these uct glycolic acid is a natural metabolite. A major appli- polymers to help the reader to further explore the use of cation of PGA is in resorbable sutures (Dexon, American these polymers or precursors with similar chemistries in Cyanamide Co). Numerous studies (Chu, 1981a,b,c) have developing injectable polymer compositions. established a simple degradation mechanism via homo- geneous erosion. The degradation process occurs in two stages, the first involves the diffusion of into the amorphous regions of the matrix and simple hydrolytic

2 P A Gunatillake & R Adhikari Polymers for tissue engineering

Table 1. Biodegradable Polyesters Polymer Polymer repeat unit structure

3 P A Gunatillake & R Adhikari Polymers for tissue engineering

Figure 1. chain scission of the ester groups. The second stage of 1984). The 70/30 GA/LA has the highest water uptake, degradation involves largely the crystalline areas of the hence the most readily degradable in the series. In an- polymer, which becomes predominant when the majority other study (Miller et al., 1977) have shown that the 50/ of the amorphous regions have been eroded. 50 was the most unstable with respect to hy- In a study of Dexon sutures in vitro, the first stage drolysis. However, it is generally accepted that interme- degradation predominates during the first 21 days and a diate copolymers are very much more unstable than the further 28 days for the degradation of the crystalline re- homopolymers. The first commercial use of this copoly- gions. After 49 days, the reported weight loss was around mer range was the suture material Vicryl (Ethicon Inc, 42 % with complete loss of mechanical properties. Be- Sommerville, NJ, USA; www.ethicon.com), which is com- cause of the bulk degradation of PGA, there is a sudden posed of 8% (l)LA and 92% GA. The main application of loss of mechanical properties. Although the degradation (d,l- LA/GA) copolymer has been in the field of control- product glycolic acid is resorbable at high concentrations, led drug release. they can cause an increase of localized acid concentra- tion resulting in tissue damage. The ultimate fate of gly- Biodegradation and biocompatibility of colic acid in-vivo is considered to be the conversion to polylactides. The degradation of PLA, PGA and PLA/ dioxide and water, with removal from the body PGA copolymers generally involves random hydrolysis via the respiratory system (Gilding, 1981). However, of their ester bonds. PLA degrades to form lactic acid Hollinger (Hollinger, 1983) has suggested that only lac- which is normally present in the body. This acid then en- tic acid follows this pathway, and that glycolic acid is ters tricarboxylic acid cycle and is excreted as water and converted into glyoxylate (by glycolate oxidase), which is carbondioxide. No significant amounts of accumulation then transferred into glycine after reacting with glycine of degradation products of PLA have been reported in transaminase. any of the vital organs (Vert et al., 1984). Carbon13 labeled Poly(lactic acid) is present in three isomeric forms d(- PLA has demonstrated little radioactivity in feces or urine ), l(+) and racimic (d,l) and the polymers are usually ab- indicating that most of the degradation products are re- breviated to indicate the chirality. Poly(l)LA and leased through respiration. It is also reported that in ad- poly(d)LA are semi-crystalline solids, with similar rates dition to hydrolysis PGA is also broken down by certain of hydrolytic degradation as PGA. PLA is more hydro- , especially those with esterase activity (William phobic than PGA, and is more resistant to hydrolytic at- and Mort, 1977). Glycolic acid also can be excreted by tack than PGA. For most applications the (l) isomer of urine. lactic acid (LA) is chosen because it is preferentially me- The rate of degradation, however is determined by fac- tabolized in the body. PLlLA, poly(lactic-glycolic acid) tors such as configurational structure, copolymer ratio, (PLGA) copolymers and PGA are among the few biode- crystallinity, molecular weight, morphology, stresses, gradable polymers with Food and Drug Administration amount of residual , porosity and site of implan- (FDA) approval for human clinical use. tation. The full range of copolymers of lactic acid and gly- Both in-vitro and in-vivo studies have been carried colic acid has been investigated. The two main series are out to ascertain the biocompatibility of PLA and PGA. those of (l)LA/GA and (dl)LA/GA. Gilding and Reed Many studies suggests that these polymers are sufficiently (1979). have shown that compositions in the 25 to 75 % biocompatible (Nelson et al. 1977, Hollinger , 1983) al- range for (l)LA/GA and 0 to 70 % for the (dl) LA/GA are though certain studies (Schakenraad et al., 1989; amorphous. For the (l)LA/GA copolymers, resistance to VanSliedregt et al., 1990, 1992; Verheyen et al., 1993) hydrolysis is more pronounced at either end of the co- suggest otherwise. Recent studies have shown that po- polymers compositions range (Miller et al., 1977; Gild- rous PLA-PGA scaffolds may be the cause of significant ing and Reed, 1979; Reed and Gilding, 1981; Vert et al., systemic or local reactions, or may promote adverse re-

4 P A Gunatillake & R Adhikari Polymers for tissue engineering sponses during the tissue repair process. PLA-PGA co- polymers used in bone repair applications have shown to be biocompatible, non-toxic and non-inflammatory (Nel- son et al., 1977; Hollinger, 1983). Since PLA-PGA have been used successfully in clinical use as sutures, their use in fixation devices or replacement implants in muscu- loskeletal tissues may be considered safe. Concerns about the biocompatibility of these materi- als have been raised when PLA and PGA produced toxic solutions probably as a result of acidic degradation (Tayler et al., 1994). This is a major concern in orthopaedic ap- plications where implants with considerable size would be required, which may result in release of degradation products with high local acid concentrations. Another concern is the release of small particles during degrada- tion, which can trigger an inflammatory response. It has Figure 2. been shown that as the material degrades the small parti- cles that break off are phagocytized by macrophages and multinucleated giant cells (Gibbons, 1992). It was also to explore this class of polymers for various applications. noted that no adverse biological responses occur espe- cially if the material volume is relatively small. In clini- Poly(propylene fumarates) cal studies where PGA was used as fracture fixation, for- Recently, polyesters based on fumaric acid have received eign-body responses or osteolytic reactions have been re- attention in the development of degradable polymers, and ported (Böstman, 1991, 1992; Böstman et al., 1992a,b). the most widely investigated is the copolyester poly(propylene fumarate) (PPF) (Figure 2). The degrada- Polylactones tion of this copolymer leads to fumaric acid, a naturally Poly(caprolactone) (PCL) is the most widely studied in occurring substance, found in the tri- cy- this family (Holland and Tighe 1992; Hayashi, 1994). PCL cle (Krebs cycle), and 1,2-propanediol, which is a com- is a semicrystalline polymer with a tem- monly used diluent in drug formulations. The copolymer perature of about –60ºC. The polymer has a low melting also has unsaturated sites in its backbone, which could be temperature (59 to 64ºC) and is compatible with a range used in subsequent cross-linking reactions. of other polymers. PCL degrades at a much lower rate PPF based degradable polymer compositions includ- that PLA and is a useful base polymer for developing long- ing injectable biodegradable materials have been reported term, implantable drug delivery systems. in the literature (Kharas et al., 1997; Peter et al., 1998a,b; Pol(caprolactone) is prepared by the ring-opening po- Temenoff and Mikos, 2000). Injectable systems developed lymerization of the cyclic monomer ε-caprolactone. Cata- based on PPF have the advantage of employing chemical lysts such as stannous octoate are used to catalyse the cross-linking overcoming some of the disadvantages in polymerization and low molecular weights can photo cross-linkable systems. Photo-cross-linkable systems be used as initiator which also can be used to control the have limited applications for treatment of deep crevices molecular weight of the polymer (In’t Veld et al., 1997; in bone. A number of studies have reported on the syn- Storey and Taylor, 1998). thesis, properties (Kharas et al., 1997; Peter et al., 1998a; Temenoff and Mikos, 2000) and in-vivo degradation Biodegradation and biocompatibility of (Yaszemski et al., 1994; Frazier et al., 1997; Peter et al., polylactones. The homopolymer has a degradation time 1998a) characteristics of poly(propylene fumarate). The of the order of two to three years (Kronenthal, 1975; Hol- copolymers degrade to propylene glycol, poly(acrylic acid- land and Tighe 1992; Middleton and Tipton, 2000). PCL co-fumaric acid) and fumaric acid (Temenoff and Mikos, with an initial average molecular weight of 50,000 takes 2000). Cross-linking usually occurs with about three years for complete degradation in-vitro methylmethacrylate or N-vinyl pyrolidone and benzoyl (Gabelnick, 1983). The rate of hydrolysis can be altered peroxide as the initiator. by copolymerisation with other lactones, for example a A number of methods have been reported to prepare copolymer of caprolactone and valerolactone degrades PPF, and each results in different polymer properties (Pe- more readily (Pitt et al., 1981). Copolymers of ε- ter et al., 1997a,b, 1999; Temenoff and Mikos 2000). Prod- caprolactone with dl-lactide have been synthesized to yield ucts with complex structure are obtained due to side reac- materials with more rapid degradation rates (e.g., a com- tions involving different modes of addition. In one method mercial suture MONOCRYL, Ethicon) (Middleton and diethyl fumarate and propylene glycol with para-toluene Tipton, 2000). PCL is considered a non-toxic and a tissue sulfonic acid catalyst are reacted at 250ºC (Sanderson, compatible material (Kronenthal, 1975). 1988). The yield in this process is only 35 %. In another Blends with other polymers and block copolymers and method, propylene glycol and fumaric acid are heated low molecular weight polyols and macromers based on initially at 145ºC and gradually increasing the tempera- caprolactone backbone are a few of the possible strategies ture to 180ºC. Poly(propylene fumarate) diol with mo-

5 P A Gunatillake & R Adhikari Polymers for tissue engineering

Figure 3. lecular weights in the range 500 to 1200 and polydispersity dergoes bulk degradation and degradation time is depend- 3 to 4 can be typically prepared by this method (Gerhart ent on polymer structure as well as other components. PPF and Hayes, 1989). A third method involves preparing the degrades by hydrolysis to fumaric acid and propylene gly- bis-(hydroxylpropyl) fumarate trimer and propylene col. Based on in-vitro studies, the time required to reach bis(hydrogen maleate) trimer by reacting propylene gly- 20% loss in original weight ranged from 84 (PPF/ß-TCP col/fumaric acid, and maleic anhydride/propylene glycol, composite) to over 200 days (PPF/CaSO4 composite) (Pe- respectively (Domb, 1989). The two trimers are then re- ter et al., 1997b; Kharas et al., 1997). ß-TCP in these com- acted at 180ºC to produce PPF. The bis-(hydroxypropyl) positions not only increased mechanical strength, but also fumarate trimer can also be prepared at ambient tempera- acts as a buffer making the pH change minimal during the ture by reacting fumaryl chloride and propylene glycol degradation process. (Peter et al., 1997a). The purified trimer is reacted at PPF does not exhibit a deleterious long-term inflam- 160ºC in the presence of transesterification catalyst anti- matory response when implanted subcutaneously in rats. mony trioxide to produce PPF. PPF with molecular weights A mild inflammatory response was observed initially and in the range 750 to 1500 could be prepared by this method. a fibrous capsule formed around the implant at 12 weeks The polydispersity ranged from 1.7 to 3. (Peter et al., 1998a). It appears that achieving high molecular weight PPF is difficult because of side reactions, particularly due to Polyanhydrides the presence of the backbone double bond. Accordingly, Polyanhydrides are one of the most extensively studied incorporation of fillers, or further reactions to form cross- (Holland and Tighe, 1992; Kohn and Langer, 1997; linked net works would be required to achieve good me- Ashammakhi and Krokkanen, 1997) classes of biodegrad- chanical strength. The mechanical properties vary greatly able polymers with demonstrated biocompatibility and depending on the method of synthesis and the cross-link- excellent controlled release characteristics. Polyanhydrides ing agent used. Mechanical properties could be improved degrades by surface erosion (Kohn and Langer, 1997) and by incorporating ceramic materials such as tricalcium their main applications are in controlled drug delivery. phosphate (TCP), calcium carbonate or calcium sulfate. Polyanhydride based drug delivery systems have been uti- These composite materials exhibit compressive strengths lized clinically (Brem et al., 1995). in the range 2 to 30 MPa. ß-TCP was particularly useful Polyanhydrides are synthesized (Figure 3) by dehy- for reinforcement, and compositions without TCP rein- dration of the diacid or a mixture of diacids by melt forcement disintegrated very early in the implant(Temenoff polycondensation (Domb and Langer, 1987). The and Mikos, 2000). dicarboxylic acid are converted to the mixed Cross-linking characteristics reported for PPF, N-vi- anhydride of acetic acid by reflux in excess acetic anhy- nyl pyrrolidone (N-VP), benzoyl peroxide, sodium chlo- dride. High molecular weight polymers are prepared by ride, and TCP indicate that for a range of formulations, melt-polycondensation of prepolymer in vacuum under the maximum temperature varied within 38 to about 48ºC, nitrogen sweep. compared to 94ºC observed for polymethylmethacrylate Langer and coworkers (Brem, et al., 1995; Burkoth (PMMA) cements. The curing times varied between 1 and and Anseth, 2000) have synthesized polyanhydrides (I) 121 min, which allows the composites to be tailored to for drug delivery applications. Polyanhydride (I) is used specific applications. The compressive strengths varied to deliver carmustine, an anticancer drug, to sites in the between 1 and 12 MPa (Peter et al., 1998a). brain where a tumor has been removed. The degradation Biocompatibility and biodegradation of PPF. PPF un- products of (I) are non-toxic and have controlled surface

(I)

6 P A Gunatillake & R Adhikari Polymers for tissue engineering

Figure 4. erosion degradation mechanism that allows delivery of Biocompatibility and biodegradation of polyanhydrides. drugs at a known rate. Polyanhydrides are biocompatible (Laurencin et al., 1990), Polyanhydrides have limited mechanical properties that have well-defined degradation characteristics, and have restrict their use in load–bearing applications such as in been used clinically in drug delivery systems (Leong et orthopaedics. For example poly[1,6-bis(carboxyphenoxy) al., 1985). Polyanhydrides degrade by hydrolysis of the hexane] has a Young’s modulus of 1.3 MPa (Leong et al., anhydride linkage. The hydrolytic degradation rates can 1985; Uhrich et al., 1997) which is well below the modulus be altered by simple changes in the polymer backbone of human bone (40 to 60 MPa). To combine good mechani- structure by choosing the appropriate diacid monomers. cal properties of polyimides with surface-eroding charac- Poly(sebasic acid) degrades quickly (about 54 days in sa- teristics of polanhydrides, poly(anhydrides-co-imides) line), while poly(1,6-bis(-p-carboxyphenoxy)hexane de- have been developed (Attawia et al., 1995; Uhrich et al., grade much more slowly (estimated 1 year). Accordingly, 1995), particularly for orthopaedic applications. Examples combinations of different amounts of these monomers include poly-[trimellitylimidoglycine-co-bis(carbox- would result in polymer with degradation properties cus- yphenoxy) hexane], and poly[pyromellitylimidoalanine-co- tom-designed for a specific application (Temenoff and 1,6-bis(carboph-enoxy)-hexane] (Attawia et al., 1995; Seidel Mikos, 2000). et al., 1996). These poly(anhydride-co-imides) have sig- Minimal inflammatory responses to sebacic acid/1,3- nificantly improved mechanical properties, particularly bis(p-carboxyphenoxy) propane (SA/CPP) systems have compressive strengths. Materials with compressive been reported when implanted subcutaneously in rats up strengths in the 50 to 60 MPa range have been reported to 28 weeks. Loose vascularized tissue had grown into for poly(anhydrides-co-imides) based on succinic acid the implant at 28 weeks, with no evidence of fibrous cap- trimellitylimidoglycine and trimellitylimidoalanine (Uhrich sule formation (Laurencin et al., 1990). No data have been et al., 1995). The degradation of these copolymers occurred reported about polymer sterilizability and heat genera- via hydrolysis of anhydride bonds, followed by the hy- tion during polymerization. A 12 week study using 2-3 drolysis of imide bonds. mm diameter full thickness defect in the distal femur of Photo cross-linkable polyanhydrides have also been rabbits showed good tolerance of the SA/CPP polymer developed for use in orthopaedic applications, particu- system and osseous tissue in the outer zone of some im- larly focusing on achieving high mechanical strength. The plants (Laurencin et al., 1990). systems developed are based on dimethacrylated anhy- drides (Muggli et al., 1998; Burkoth and Anseth, 2000). Tyrosine-derived polycarbonates Figure 4 shows dimethacrylate macrommers based on Tyrosine-based polycarbonates (Figure 5) have been re- sebacic acid and 1,6-bis(p-carboxyphenoxy)hexane. Both ported as promising degradable polymers for use in or- ultraviolet (UV) and visible light cure methods have been thopaedic applications (Muggli et al., 1998; investigated with these macromonomers. The most effec- Tangpasuthadol et al., 2000a,b). These polymers possess tive means of photopolymerization of these three potentially hydrolysable bonds: , carbonate and macromonomers was found to be 1.0 wt % ester. Studies have shown (Muggli et al., 1998) that the camphorquinone and 1.0 wt % ethyl-4-N,N-dimethyl carbonate group hydrolyzes at a faster rate than the ester aminobenzoate with 150 mW/cm2. Combination of group, and the amide bond is not labile in vitro. Since the type and visible initiation has provided means of achiev- hydrolysis of the carbonate groups yields two alcohols and ing efficient curing of thick samples. , the problem of acid bursting seen in poly- Depending on the monomers used, the mechanical is alleviated. By variation of the structure of the properties as well as degradation time can be varied. pendant R group, polymers with different mechanical Compressive strengths of 30-40 MPa, and tensile strengths properties, degradation rates as well as cellular response of 15-27 MPa, similar to those of cancelleous bone, have could be prepared. Polycarbonate having an ethyl ester been reported (Anseth et al., 1997). pendant group has shown to be strongly osteoconductive

7 P A Gunatillake & R Adhikari Polymers for tissue engineering

Figure 5. and good bone apposition, and possesses sufficient me- in the fabrication of medical implants such as cardiac pace chanical properties for load bearing bone fixations. In- makers and vascular grafts. Recent developments in vivo studies have demonstrated that the polymer was siloxane-based polyurethanes, which have greater in-vivo biocompatible and promoted significant bone growth stability than conventional polyetherurethanes (e.g., (Pulapura and Kohn, 1992; Muggli et al., 1998). poly(tetramethylene oxide) (PTMO)-based) have provided opportunities for development of a range of medical im- Polyorthoesters plants for chronic applications (Gunatillake et al., 2001). Poly(orthoester)s (POE) are another family of polymers Polyurethanes can also be designed to have chemical identified as degradable polymers suitable for orthopae- linkages that are degradable in the biological environ- dic applications. Heller and coworkers reported on the ment (Zdrahala and Zdrahala, 1999). Since polyurethanes synthesis of a family of polyorthoesters (Figure 6) that can be tailored to have a broad range of mechanical prop- degrades by surface erosion (Ng et al., 1997). With the erties and good biocompatibility, there has been some in- addition of lactide segments as part of the polymer struc- terest to develop degradable polyurethanes for medical ture, tunable degradation times ranging from 15 to hun- applications such as scaffolds for tissue engineering dreds of days can be achieved. The degradation of the (Zdrahala and Zdrahala, 1999). However, a major prob- lactide segments produces carboxylic acids, which catalyze lem has been the of degradation products, par- the degradation of the orthoester (Ng et al., 1997). ticularly those derived from the diisocyanate component. Preliminary in-vivo studies have shown that POE (Fig- For example, degradation products of polyurethanes based ure 6) to increase bone growth in comparison with poly(di- on diisocyanates such as 4,4’-methylenediphenyl lactide-co-glycolide) (Andriano et al., 1999). diisocyanae (MDI) and toluene diisocyanate (TDI) are toxic (McGill and Motto, 1974; Gogolewski and Pennings, Polyurethanes 1982). Accordingly, in designing degradable poly- Polyurethanes (PU) represent a major class of synthetic urethanes diiscoyanates such as lysine diisocyanate (LDI) elastomers that have been evaluated for a variety of medi- (2,6-diisocyanatohexanoate) and other aliphatic cal implants, particularly for long-term implants (Pinchuk, diisocyanates like hexamethylene diisocyanate (HDI) and 1994; Lamba et al., 1998). They have excellent mechani- 1,4-butanediisocyanate have been used. cal properties and good biocompatibility. They are used

Figure 6.

8 P A Gunatillake & R Adhikari Polymers for tissue engineering

Properties and bidegradation of aliphatic diisocyanate in in-vitro. Subcutaneous implantation in guinea pigs base polyurethanes: showed that the porous networks allowed rapid cell in-growth, degraded almost completely 4-8 A number of studies have been reported on the synthesis weeks after implantation and evoked no adverse tissue and properties of a range of polyurethanes based on lysine reaction. diisocyanate (LDI) (II). Zang et al. (2000) have developed a peptide based poly- urethane scaffold for tissue engineering. LDI was reacted first with glycerol to form a prepolymer, which upon re- action with water produced a cross-linked porous sponge due to liberation of carbon dioxide. Initial cell growth studies with rabbit bone marrow stromal cells have shown (II) that the polymer matrix supported cell growth and was phenotypically similar to those grown on tissue culture polystyrene. Hirt et al. (1996) and De Groot et al. (1990) reported on the synthesis and properties of degradable poly- urethanes based on LDI, 2,2,4-triethylhexamethylene diisocyanate and a number of polyester and copolyester (III) polyols such as Diorez®, caprolactone, ethylene glycol copolymers, and poly hydroxy butyrate and valerate co- polymers. The polyurethanes ranged from elastomers with elongations at break as high as 780 %, but with low ten- Lysine diisocyanate is not commercially available (be- sile strengths (5.8 to 8.1 MPa). Saad et al. (1997) re- ing developed by Kyowa Hakko Kogyo Co., Chiyoda-Ku, ported on the cell and tissue interaction of four such poly- Tokyo, Japan) but can be prepared from L-lysine mers prepared from 2,2,4-trimethylhexamethylene monohydrochloride (Bruin et al., 1988; Storey et al., diisocyanate and 2,6-diisocyanato methyl caproate, and 1993). Storey et al (1994) have prepared poly(ester polyols a,ω-dihydroxy-poly(R-3-hydroxybutyrate-co-(R)- urethane) networks from LDI and a series of polyester 3-hydroxyvalerate)-block-ethylene glycol], and two com- triols based on dl-lactide, γ-caprolactone, and their co- mercial diols, Diorez® and PCL-diol. In-vitro studies in- polymers. Networks based on poly(dl-lactide) were rigid dicated that these polyesterurethanes did not activate (glass transition temperature Tg = 60ºC) with ultimate macrophages and showed good level of cell adhesion, and tensile strengths of ~ 40 to 70 MPa, whereas those based growth, which were also confirmed by in-vivo results. on caprolactone triols were low modulus elastomers with Structure-property relationships of degradable poly- tensile strengths of 1 to 4 MPa. Networks based on co- urethanes based on 2,6-diisocyanato methyl caproate, polymers were more elastomeric (elongation up to 600%) , oxide (PEO) and an amino with compressive strengths between 3 to 25 MPa. Hydro- acid chain extender (phenylalanine) have been investi- lytic degradation under simulated physiological conditions gated by Skarja and Woodhouse (1998, 2000). Their re- were dependent on the type of triol and dl-lactide based sults showed that PEO based polyetherurethane (PEU) networks were the most resistant with no degradation were generally weaker but PCL based materials were rela- observed for 60 days, caprolactone based triol networks tively strong. However, no results were reported on the were resistant up to 40 days whereas the high lactide based degradation of these polyurethanes. copolymer networks were the least resistant and substan- Gogolewsky and Pennings (1982, 1983) have reported tial degradation observed in about 3 days. on a design of an artificial skin composed of polylactide/ Bruin et al. (1988) have reported on the synthesis of polyurethane mixtures where the PU was non-degrada- degradable polyurethane networks based on star-shaped ble. In-vivo studies with guinea pigs showed that the arti- polyester prepolymers. The star-prepolymers were pre- ficial skin adhered to wound well, and protected from fluid pared from myoinisitol, a pentahydroxy sugar molecule loss and infections up to 40 days exhibiting potential as a by ring-opening copolymerisation of l-Lactide or glycolide skin substitute. with caprolactone. The prepolymers were cross- linked Micro-porous polyurethane amide and polyurethane- using 2,6-diisocyanatohexanoate. The degradation prod- urea scaffolds have been evaluated by Spaans et al. (2000) ucts of these PU networks are considered non-toxic. The for repair and replacement of knee-joint meniscus. The resulting network polymers were elastomeric with elon- soft segments in these polyurethanes were based on 50/ gation in the range 300 to 500% and tensile strengths 50 l-lactide/PCL and chain extenders were adipic acid varying between 8 to 40 MPa depending on the branch and water, the reaction of latter with 1,4-butane length etc. Preliminary experiments in guinea pigs have diisocyanate provided carbon dioxide to produce porous shown that the polyurethanes biodegrade when implanted scaffolds. Salt crystals were also added to produce porous subcutaneously. Polyurethane networks based on LDI and structure, and the addition of surfactants combined with poly(glycolide-co-γ-caprolactone) macrodiol was evalu- ultrasonic waves regulated the pore structure. Porous scaf- ated by Bruin et al. (1990) as two-layer artificial skin. folds with porosity of 70 to 80% were achieved by this The degradation of the skin in-vivo was faster than that technique. These scaffolds exhibited tearing problems

9 P A Gunatillake & R Adhikari Polymers for tissue engineering during suturing (De Groot et al.., 1996), which was partly tissue engineering applications requires the preparation of circumvented by using a different suturing system. A precursors with appropriate physical properties and func- meniscal replica implanted contained only fibro-cartilage tional groups for curing at a second stage. A choice of a after 18 weeks and decreased the degradation of the ar- suitable curing method with minimal heat generation and ticular cartilage. chemical reactions that do not interfere with biological com- Biocompatibility and biodegradation of degradable ponents is also very important for developing such poly- polyurethanes. Although a number of studies discussed mer compositions. A range of oligomeric precursors with above indicate that the biocompatibility of degradable degradable backbones has been reported in the literature polyurethanes appear to be satisfactory based on both in- (Leong et al., 1985, Attawia et al., 1995; Uhrich et al., 1995; vitro and in-vivo studies. Animal studies showed rapid Seidel et al., 1996; Kharas et al., 1997; Uhrich et al., 1997; cell in-growth with no adverse tissue reactions. However, Muggli et al., 1998; Peter et al., 1998b; Burkoth and Anseth, the effect of degradation products and how those prod- 2000; Temenoff and Mikos, 2000) with potential to develop ucts are removed from the body are not clearly under- such polymer compositions. Many of these are based on stood. the various families of degradable polymers discussed pre- viously. Majority of them contain ester functional groups Polyphosphazenes in the backbone. Table 2 provides a summary of the prop- The polyphosphazenes consist of several hundred differ- erties of common biodegradable polymers while Table 3 ent polymers with the general structure (VI) (Mark et al., provides some of the macrodiols and macromers with de- 1992). Different polyphosphazenes are made by means of gradable backbones suitable for the development of inject- macromolecular substitution reactions carried out on a able polymer compositions. reactive polymeric intermediate, poly(dichloro- Only a few synthetic polymer-based injectable compo- phosphazene), (NPCl2)n. Although most polypho- sitions have been reported in the literature. Poly(propylene sphazenes are biostable, incorporation of specific side fumarate) (Kharas et al., 1997; Peter et al., 1998b; Temenoff groups such as esters, glucosyl, glyceyl, lac- and Mikos, 2000) and dimethacrylated polyanhydrides tate, or imidazolyl units can render polyphosphazenes (Leong et al., 1985; Attawia et al., 1995; Uhrich et al., 1995; biodegradable (Alcock, 1999; Behravesh et al., 1999; Qui Seidel et al., 1996; Uhrich et al., 1997, Muggli et al; 1998; and Zhu, 2000). Burkoth and Anseth, 2000) are two types of precursors that have been reported recently as in-situ polymerizable sys- tems with potential for orthopaedic tissue engineering ap- plications. Poly(ortho esters) based injectable polymer has also been reported (Heller et al., 2002) for use in pain con- trol and periodontal treatment. The in-situ cross-linking of fumarates has been achieved by using benzoyl perox- ide initiator in the presence of methylmethacrylate and N- (VI) vinyl pyrolidone while UV and visible light initiators have been used for dimethacrylated polyanhydrides.

Hydrolysis of the polymer leads to free side group units, Conclusions phosphate and ammonia due to backbone degradation. Biocomaptibility and biodegradation. Laurencin et A vast majority of biodegradable polymers studied be- al. (1993) have investigated methylphenoxy and either long to polyester family and poly(glycolic acid), poly (lactic imidazolyl or ethylglycinate substituted polyphosphazenes acid) and their copolymers have historically comprised for skeletal tissue regeneration. Both materials supported the bulk of published material. These polymers have a the growth of MC3T3-E1, an osteogenic cell line. Increase relatively long history of use in a number of clinical ap- in imidazolyl side groups resulted in a reduction in cell plications. They will continue to play a key role in vari- attachment and growth on the polymer surface and an ous forms for medical applications requiring biodegrad- increase in the rate of degradation of the polymer. In con- able polymers. Polyesters offer synthetic chemists many trast, substitution with ethylglycinato group favoured in- opportunities to design polymers through combination of creased cell adhesion and growth and also an increase in different monomers to achieve property requirements to the rate of degradation of the polymers. suit a variety of applications. Additionally, the develop- In another study (Laurencin et al., 1996), porous ma- ment of precursors such as polyols and macromonomers trices of poly[(50% ethylglycinato) (50% p- based on polyesters may find uses in injectable and in- methylphenoxy) phosphazene] with pore sizes of 150 to situ curable polymer formulations. Poly(propylene fuma- 250 µm have been shown to be good substrate for osteob- rate) is one example of a recently developed polyester- last-like cell attachment and growth. based injectable polymer system. Polyanhydride is another family of polymers studies Development of injectable and biodegradable extensively with demonstrated biocompatibility and ex- polymers for tissue engineering cellent controlled release characteristics. Polyanhydride degrade by bulk erosion and their main applications are in The devlopment of injectable polymer compositions for controlled drug delivery. Recently photocross-linkable

10 P A Gunatillake & R Adhikari Polymers for tissue engineering

Table 2 Properties of biodegradable polymers

Polymer Thermal & Mechanical Properties Melting Glass Approximate Processing Approx. Degration Degradation Biocompatibility and Point Transition Strength Method Time (months) Products Biodegradation (ºC) (ºC) (References)

Polyesters

Poly(glycolic 225-230 35-40 7.0 GPa E, IM, CM, SC 6 to 12 Glycolic acid Many studies have shown that acid) (Modulus) polyglycolides, polylactides and their copolymers to have Poly(l-lactic 173-178 60-65 2.7 GPa E, IM, CM, SC >24 l-lactic acid acceptable biocompatibility. acid) (Modulus) Some studies have shown systemic or local reactions due Poly(d,l- amorphous 55 to 60 1.9 GPa E, IM, CM, SC 12 to 16 d,l-lactic acid to acidic degradation products. lactic acid) (Modulus) Biodegradation of these polymers takes place by random Poly(d,l-lactic- amorphous 50 to 55 2.0 GPa E, IM, CM, SC 5 to 6 d,l-lactic acid hydrolysis resulting in decrease co-glycolic (Modulus) and in molecular weight followed, acid) (85/15) glycolic acid first, by a reduction in mechanical properties and mass Poly(d,l-lactic- amorphous 50 to 55 2.0 GPa E, IM, CM, SC 4 to 5 d,l-lactic acid loss. Natural pathways co-glycolic (Modulus) and (, excretion) acid) (85/15) glycolic acid harmlessly eliminate the final degradation products. Poly(d,l-lactic- amorphous 45 to 50 2.0 GPa E, IM, CM, SC 3 to 4 d,l-lactic acid (Middleton and Tipton, 2000) co-glycolic (Modulus) and acid) (85/15) glycolic acid

Poly(d,l-lactic- amorphous 45 to 50 2.0 GPa E, IM, CM, SC 1 to 2 d,l-lactic acid co-glycolic (Modulus) and acid) (85/15) glycolic acid Generally considered as a non- Poly(capro 58 to 63 -65 to 60 0.4 GPa E, IM, CM, SC >24 Caproic acid toxic and tissue compatible lactone) (Modulus) polymer (Holland and Tighe, 1992; Hayashi, 1994) Poly(propylene 2 to 30 MPa Injectable prepolymer Depends on the Fumaric acid, Initial mild inflammatory fumarate) - - (compressive cross-linked via free formulation, several propylene glycol response and no deleterious lone- strength) radical initiation months based on in- and poly(acrylic term response based on rat vitro data acid-co-fumaric implant studies (Kharas et al. acid) 1997; Peter et al. 1998a; Temenoff and Mikos, 2000) Polyanhydrides, polycarbonates, polyurethanes, and polyphosphazenes Poly[1,6- 1.3 MPa (Youngs Polyanhydrides are - - Thermoplastic 12 (in-vitro) bis(carboxyphenoxy) modulus) Dicarboxylic biocompatible and have well hexane] acids defined degradation characteristics. Degrades by hydrolysis of the anhydride linkage (surface degradation). (Leong et al. 1985; Uhrich et al. 1997) Tyrosine- Sufficient Very slow Tyrosine, Biocompatible and promotes derived - -mechanical strength Thermoplastic degradation (in- carbondioxide bone growth (in-vivo studies) polycarbonate for load bearing vitro) and alcohols (Pulpura and Kohn, 1992; bone fixation Muggli et al., 1998) Polyurethane Lysine, glycolic No adverse tissue reaction based on LDI and - -8 to 40 MPa tensile Castable thermoset 1 to 2 and caproic acids (guinea pigs) (Bruin et al., 1988) poly(glycolide-co- strength γ-caprolactone)

Ethylglycinate - - -Thermally >1 (in-vitro) Phosphates & Biocompatible and support polyphosphazene processable ammonia from osteogenic cell growth (in-vitro) backbone and (Qui and Zhu, 2000) other products depending on side chain structure

This summary was prepared based on the information taken from the literature reviewed in this article and the data should be used as a guide only E = extrusion, IM = injection moulding, CM = compression moulding, SC = solvent casting, Sufficient data are not available to provide comprehensive list of properties for polymers in these families.

11 P A Gunatillake & R Adhikari Polymers for tissue engineering

Table 3. Precursors for developing injectable biodegradable polymers

12 P A Gunatillake & R Adhikari Polymers for tissue engineering polyanhydrides have been developed for use in orthopae- Böstman OM (1992) Intense granulomatous inflamma- dic applications. Tyrosine-derived polycarbonates, tory lesions associated with absorbale internal fixation polyorthoseters, polyurethanes and polyphosphazenes devices made of in ankle fractures. Clin have also been investigated to explore their potential as Orthop 278: 178-199. biodegradable polymers. Böstman O, Päivaärinta U, Partio E, Vasenius J, Review of the literature indicates that relatively few Manniner M, Rokkanen P (1992a) Degradation and tis- attempts have been made to develop injectable polymer sue replacement of an absorbable polyglycolide screw in compositions for use in tissue engineering applications. the fixation of rabbit osteomies. J Bone Joint Surg 74A: The key challenges in developing such compositions in- 1021-1031. clude the choice of appropriate precursors that would de- Böstman O, Partio E, Hirvensalo E, Rokannen P grade to biocompatible and resorbable compounds, the (1992b) Foreign-body reactions to polyglycolide screws. ability to incorporate cells and other components to sup- Acta Orthop Scand 63: 173-176. port cell attachment and proliferation, ability to cure in- Brem H, Piantadosi S, Burger PC, Walker M, Selker situ in the in-vivo environment with minimal heat gen- R, Vick NA, Black K, Sisti M, Brem G, Mohr G, Muller eration, and the ability to control degradation kinetics to P, Morawetz R, Schold SC (1995) Lancet 345: 1008. suit the intended application. Bruin P, Venstra GJ, Nijenhuis AJ, Pennings AJ (1988) Polyurethanes offer many advantages in the design of Design and synthesis of biodegradable poly(ester- injectable and compositions. As a urethane) elastomer networks composed of non-toxic class of polymers, polyurethanes generally have good building blocks. Makromol Chem Rapid Commun 9: 589- biocompatibility. They also offer substantial opportuni- 594. ties to tailor polymer structure to achieve a broad range of Bruin P, Smedinga J, Pennings, AJ, Jonkman MF mechanical properties. By choice of star, dendritic or (1990) Biodegradable lysine diisocyanate-based hyperbranched prepolymers, one can introduce structural poly(glycolide-co-ε-caprolactone)-urethane network in variations to tailor degradation kinetics as well as incor- artificial skin. 11: 191-295. poration of appropriate functional groups for improved Burg KJL, Porter S, Kellam JF (2000) Biomaterials cell attachment. In the rapidly advancing field of tissue development for bone tissue engineering. Biomaterials 21: engineering, polyurethanes offer numerous opportunities 2347-2359. to develop suitable scaffolds for a variety of applications. Burkoth AK, Anseth KS (2000) A review of photocrosslinked polyanhydrides: in situ forming degra- References dable networks. Biomaterials 21: 2395-2404. Chu CC (1981a) An in-vitro study of the effect of buffer Agrawal CM, Athanasiou KA, Heckman JD (1997) on the degradation of poly(glycolic acid) sutures. J Biomed Biodegradable PLA/PGA polymers for tissue engineer- Mater Res 15: 19-27. ing in orthopaedica. Material Science Forum 250: 115- Chu CC (1981b) The in-vitro degradation of 128. poly(glycolic acid) sutures- effect of pH. J Biomed Mater Alcock HR (1999) Inorganic-organic polymers as route Res 15: 795-804. to biodegradable materials. Macromol Symp 144: 33-46. Chu CC (1981c) Hydrolytic degradation of Andriano KP, Tabata Y, Ikada Y, Heller J (1999) In polyglycolic acid: tensile strength and crystallinity study. vitro and In vivo comparison of bulk and surface hydroly- J Appl Polym Sci 26: 1727-1734. sis in absorbable polymer scaffolds for tissue engineer- Cutright D, Beasley J, Perez B (1971) Histologic com- ing. J Biomed Mater Res (Appl Biomater) 48: 602-612. parison of sutures. Oral Surg 32: 165-173. Anseth KS, Svaldi DC, Laurencin CT, Langer R (1997) De Groot JH, Nijenhuis AJ, Bruin P, Pennings AJ, Photopolymerisation of novel degradable networks for Veth RPH, Klompmaker J, Jansen HWB (1990) Use of orthopaedic applications. In: Photopolymerization. porous biodegradable polymer implants in meniscus re- Scranton A, Bowman C, Peiffer R, eds. ACS Symposium construction. 1. preparation of porous biodegradable poly- Series 673; American Chemical Society, Washington DC. urethanes for the reconstruction of meniscus lesions. Col- pp 189-202. loid Polymer Sci 268: 1073-1081. Ashammakhi N, Rokkanen P (1997) Absorbable De Groot JH, De Vrijer R, Pennings AJ, Klompmaker polyglycolide devices in trauma and bone surgery. J, Veth RPH, Jansen HWB (1996) Use of porous poly- Biomaterials 18: 3-9. urethanes for meniscal reconstruction and meniscal Attawia MA, Uhrich KE, Botchwey E, Fan M, Langer prosthses Biomaterials. 17: 163-173. R, Laurencin CT (1995) Cytotoxocity testing of Domb AJ (1989) Poly(propylene glycol fumarate) com- poly(anhydride) for orthopaedic applications. J Biomed positions for biomedical applications. United States Pat- Mater Res 29: 1233-1240. ent 4888 413: 1-31. Behravesh E, Yasko AW, Engle PS, Mikos AG (1999) Domb AJ, Langer R (1987) Polyanhydrides I: Prepa- Synthetic biodegradable polymers for orthopaedic appli- ration of high molecular weight polyanhydrides. J Polym cations. Clin Orthop 367S: 118-185. Sci, Part A, Polymer Chem 25: 3373-3386. Böstman OM (1991). Osteolytic changes accompany- Frazier DD, Lathi VK, Gerhart TN, Hayes WC (1997) ing degradation of absorbable fracture fixation implants. J Ex vivo degradation of a poly(propylene glycol-fumarate) Bone Joint Surg 73B: 679-682. biodegradable particulate composite bone cement. J

13 P A Gunatillake & R Adhikari Polymers for tissue engineering

Biomed Mater Res 35: 383-389. Dijkstra PJ, Feijen J (1997) Melt block copolymerisation of Gabelnick HL (1983) Long acting steroid contracep- e-caprolactone and L-Lactide. J. Polym Sci Part 1 Polymer tion. In: Advances in Human Fertility and Reproductive Chem 35: 219-226. Endocrinology. Mishell Jr DR, ed. Raven Press, New York. Jen AC, Peter SJ, Mikos AG (1999) Preparation and Vol 3, pp. 149-173 use of porous poly(a-hydroxyester scaffolds for bone tis- Gerhart TN, Hayes WC (1989) Bioerodible implant sue engineering. In: Tissue Engineering Methods and composition. United States Patent 4843 112: 1-16. Protocols. Morhgan JR, Yarmush ML eds Humana Press, Gibbons DF (1992) Tissue response to resorbable syn- Totowa. pp 133-140. thetic polymers. In: Degradation Phenomena on Polymeric Kharas GB, Kamenetsky M, Simantirakis J, Beinlich Biomaterials. Plank H, Dauner M Renardy M, eds. KC, Rizzo AT, Caywood GA, Watson K (1997) Synthesis Springer Verlag, New York. pp 97-104. and characterization of fumarate-based polyester for use Gilding DK (1981) In: Biodegradable Polymers. in bioresorbable bone cement composites. J Appl Poly- Biocompatibility of Clinical Implant Materials Vol II. mer Sci 66: 1123-1137. Williams DF, ed. CRC Press, Boca Raton, FL. Vol 2, pp Kohn J, Langer R (1997) Bioresorbable and bioerodible 209-232. materials. In: An Introduction to Materials in Medicine. Gilding DK, Reed AM (1979) Biodegradable poly- Ratner BD, Hoffman AS, Schoen FJ, Lemon JE, eds. Aca- mers for use in surgery-polyglycolic/poly(acetic acid) demic Press, San Diego. pp 65-73. homo- and copolymers: 1. Polymer 20: 1459-1464. Kronenthal RL (1975) Biodegradable polymers in Gogolewski S, Pennings AJ (1982) Biodegradable medicine and surgery. Polymer Sci Technol 8: 119-137. materials of polylactides, 4 Porous biomedical materials Kulkarni RK, Pani KC, Neuman C, Leonard F(1966) based on mixtures of polylactides and polyurethanes. Polylactic acid for surgical implants. Arch Surg 93: 839- Makromol Chem Rapid Commun 3: 839-845. 843. Gogolewski S, Pennings AJ (1983) An artificial skin Lamba NMK, Woodhouse KA, Cooper SL (1998) Poly- based on biodegradable mixtures of polylactides and poly- urethanes in Biomedical Applications. CRC Press: New urethanes for full-thickness skin wound covering. York p 205-241. Makromol Chem Rapid Commun 4: 675-680. Laurencin CT, Peirrie-Jacques HM, Langer R (1990) Gunatillake PA, Meijs GF, McCarthy SJ (2001) De- Toxicology and biocompatibility considerations in the velopments in design and synthesis of biostable poly- evaluation of polymeric materials for biomedical appli- urethanes. In: Biomedical Applications of Polyurethanes. cations. Clin Lab Med 10: 549-570. Vermette P, Griesser HJ, Laroche, G, Guidoin R, eds. Laurencin CT, Maria EN, Elgendy HM, El-Amin SF, Landes Bioscience, Georgetown. pp 160-170. Allcock HR, Pucher SW, Ambrosio AA(1993) Use of Hayashi T (1994). Biodegradable polymers for bio- polyphosphazenes for skeletal tissue regeneration. J medical applications. Prog Polymer Sci 19: 663-702. Biomed Mater Res 27: 963-973. Heller J, Barr J, Ng SY, Shen H-R, Schwach- Laurencin, CT, El-Amin SF, Ibim, SE, Willoughby Abdellaoui K, Gurny R, Vivien-Castioni N, Loup PJ, DA, Attawia M, Allcock HR, Ambrosio AA (1996) A Baehni P, Mombelli A (2002). Development and applica- highly porous 3-dimentional polyphophazene polymer tions of injectable poly(ortho esters) for pain control and matrix for skeletal tissue regeneration. J Biomed Mater periodontal treatment. Biomaterials 23: 4397-4404. Res 30: 133-138. Hirt TD, Neuenschwander P, Suter UW (1996) Syn- Leong KW, Brott BC, Langer R (1985) Bioerodible thesis of degradable, biocompatible, and tough block- polyanhydrides as drug-carrier matrices. I: Characteriza- copolyesterurethanes. Macromol Chem Phys 197: 4253- tion, degradation, and release characteristics. J Biomed 4268. Mater Res 19: 941-955. Hofmann GO (1995) Biodegradable implants in trau- Mark JE, Alcock HR, West R (1992) Inorganic Poly- matology: A review on the state-of-the-art. Arch Orthop mers. Prentice Hall, Upper Saddle River, NJ. Chapter 3. Trauma Surg 114: 123-132. Mayer MH, Hollinger JO (1995) Biodegradable bone Holland SJ, Tighe BJ (1992) Biodegradable polymers. fixation devices. In: Biomedical Applications of Synthetic In: Advances in Pharmaceutical Science. Academic Press, Biodegradable Polymers. Hollinger JO, ed. CRC Press, London. Vol 6, pp 101-164. Boca Raton, FL. pp 173-195. Hollinger JO (1983) Preliminary report on osteogenic McGill DB, Motto JD (1974) An industrial outbreak potential of a biodegradable copolymer of polylactide of toxic hepatitis due to methylenedianiline. New Engl J (PLA) and polyglycolide (PGA). J Biomed Mater Res 17: Med 291: 278-282. 71-82. Middleton JC, Tipton AJ (2000) Synthetic biodegradable Hollinger JO, Jamiolkoski DD, Shalaby SW (1997) polymers as orthopaedic devices. Biomaterials 21: 2335- Bone repair and a unique class of bidegradable polymers: 2346. The polyesters. In: Biomedical Applications of Synthetic Mikos AG, Temenoff JS (2000) Formation of highly Biodegradable Polymers. Hollinger JO, ed. CRC Press, porous biodegradable scaffolds for tissue engineering. EJB Boca Raton, FL. pp 197-222. Electron J Biotechnol 3: 114-119 (available on http:// Hubbell J (1995) Biomaterials in tissue engineering. www.ejb.org/content/vol3/issue2/full/5). Biotechnology 13: 565-576. Miller RA, Brady JM, Cutright DE (1977) Degradation In’tVeld PJA, Velner EM, Van DeWhite P, Hamhuis J, rates of oral resorbable implants (polylactates and

14 P A Gunatillake & R Adhikari Polymers for tissue engineering polyglycolates): rate modification with changes in PLA/ 74. PGA copolymer ratios. J Biomed Mater Res 11: 711-719. Sanderson JE (1988) Bone replacement and repair Muggli DS, Burkoth AK, Keyser SA, Lee HR, Anseth putty material from unsaturated polyester resin and vinyl KS (1998) Reaction behaviour of biodegradable, photo pyrolidone. United States Patent 4,722948: 1-14. cross-linkable polyanhydrides. Macromolecules 31: 4120- Schakenraad JM, Nieuwenhues P, Molenaar I, Helder 4125. J, Dykstra PJ, Feijen J (1989) In-vivo and in-vitro degra- Nelson JF, Stanford HG, Cutright DE (1977) Evalua- dation of glycine/DL-lactic acid copolymers. J Biomed tion and comparison of biodegradable substances as os- Mater Res 23: 1271-1288. teogenic agents. Oral Surg 43: 836-843. Seidel JO, Uhrich KE, Laurencin CT, Langer R (1996). Ng SY, Vandamme T, Tayler MS, Heller J (1997) Syn- Erosion of poly(anhydride-co-imides): A preliminary thesis and erosion studies of self-catalysed mechanistic study. J Appl Polym Sci 62: 1277-1283. poly(orthoester)s. Macromolecules 30: 770-772. Shalaby SW (1988) Bioabsorbable Polymers. In: En- Peter SJ, Miller MJ, Yaszemski MJ, Mikos AG (1997a) cyclopedia of Pharmceutical Technology. Swarbrick J, Poly(propylene fumarate). In: Handbook of Biodegrad- Boylan JC, eds. vol 1, pp 465-476. able Polymers. Domb AJ, Kost J, Wiseman DM eds. Skarja GA, Woodhouse. KA (1998) Synthesis and Harwood Academic Publishers, Amsterdam. pp 87-98. characterization of degradable polyurethane elastomers Peter SJ, Nolley JA, Widmer MS, Merwin JE, containing an amino acid-based chain extender. J Yazemski MJ, Yasko AW, Engel PS, Mikos AG (1997b) Biomater Sci Polym Ed 9: 271-295. In vitro degradation of a poly(propylene fumarate)/ß- Skarja GA, Woodhouse KA (2000) Structure-property tricalcium phosphate composition orthopaedic scaffold. relationships of degradable polyurethane elastomers con- Tissue Eng 3: 207-215. taining an amino acid-based chain extender. J Appl Poly- Peter SJ, Miller ST, Zhu G, Yasko AW, Mikos AG mer Sci 75: 1522-1534. (1998a) In vivo degradation of a poly(propylene fuma- Spaans CJ, Belgraver VW, Rienstra O, De Groot JH, rate)/ß-tricalcium phosphate injectable composite scaffold. Veth RPH, Pennings AJ(2000) Solvent-free fabrication J Biomed Mater Res 41: 1-7. of micro-porus polyurethane amide and polyurethane-urea Peter SJ, Miller MJ, Yasko AW, Yaszemski MJ, Mikos scaffolds for repair and replacement of the knee-joint AG (1998b) Polymer concepts in tissue engineering. J meniscus. Biomaterials 21: 2453-2460. Biomed Mater Res (Appl Biomater) 43: 422-427. Storey RF, Taylor AE (1998). Effect of stannous octoate Peter SJ, Suggs AJ, Yazemski MJ, Engle PS, Mikos on the composition, molecular weight, and molecular AG (1999) Synthesis of poly(propylene fumarate) by acyla- weight distribution of ethyleneglycol-initiated poly(e- tion of propylene glycol in the presence of a proton scav- caprolactone). J Macromol Sci- Pure Appl Chem A35: enger. J Biomater Sci Polym Edn 10: 363-373. 723-750. Pitt CG, Marks TA, Schindler A. (1981) Naltrexone. Storey RF, Wiggins JS, Mauritz KA, Puckett AD Willette RE, Barnet G, eds National Institute on Drug (1993) Bioabsorbable composites. II: Nontoxic, L-lysine- Abuse Research Monograph. 28: 232-253. based (polyester-urethane) matrix composites. Polymer Pinchuk L (1994) A review of the biostability and Composites 14: 17. carcinogenicity of polyurethanes in medicine and the new Storey RF, Wiggins JS, Puckett AD (1994) generation of ‘biostable polyurethanes. J Biomater Sci Hydrolyzable poly(ester-urethane) networks from l-lysine Polymer Edn 6: 225-267. diisocyanate and d,l-lactide/e-caprolactone homo and Pulapura S, Kohn J (1992) Tyrosine derived copolyester triols. J Polymer Sci: Part A: Polymer Chem- polycarbonates: backbone modified “pseudo-poly(amino istry 32: 2345-2363. acids) designed for biomedical applications. Biopolymers Tangpasuthadol V, Pendharkar SM, Kohn J (2000a) 32: 411-417. Hydrolytic degradation of tyrosine-derived polycarbonates, Qui LY, Zhu KJ (2000). Novel biodegrable a class of new biomaterials. Part I: Study of model com- polyphosphazenes containing glycine ethyl ester and ben- pounds. Biomaterials 21: 2371-2378. zyl ester of amino acethydroxamic acid as cosubsituents: Tangpasuthadol V, Pendharkar SM, Peterson RC, synthesis, characterization and degradation properties. J Kohn J (2000b) Hydrolytic degradation of tyrosine-de- Appl Polym Sci 77: 2987-2955. rived polycarbonates, a class of new biomaterials. Part II: Reed AM, Gilding DK (1981) Biodegradable poly- Study of model compounds. Biomaterials 21: 2379-2387. mers for use surgery-poly(glycolic)/polylactic acid) homeo Tayler MS, Daniels AU, Andriano KP, Heller J (1994) and copolymers 2. In-vitro degradation. Polymer 22: 494- Six bioabsorbale polymers: In-vitro acute toxicity of ac- 504. cumulated degradation products. J Appl Biomater 5: 151- Sawhney AS, Drumheller PD (1998) Polymer synthe- 157. sis. In: Frontiers in Tissue Engineering. Patrick CW, Temenoff JS, Mikos AG (2000) Injectable biodegrad- Mikos AG, McIntire LV, eds. Pergamon, New York. pp able materials for orthopaedic tissue engineering. 83-106. Biomaterials 21: 2405-2412. Saad B, Hirt TD, Welti M, Uhlschmid GK, Thomson RC, Wake MC, Yaszemski, Mikos AG Neuenschwander P, Suter UW (1997) Development of (1995a) Biodegradable polymer scaffolds to regenerate degradable polyesterurethanes for medical applications: In organs. Adv Polymer Sci 122: 245-274. vitro and in vivo evaluation. J Biomed Mater Res 36: 65- Thomson RC, Wake MC, Yaszemski MJ, Mikos AG

15 P A Gunatillake & R Adhikari Polymers for tissue engineering

(1995b) Biodegradable polymer scaffolds to engineer Biomaterials 21: 1247-1258. trabecular bone. J Biomater Sci Polymer Ed 7: 23-38. Zdrahala RJ, Zdrahala IJ (1999) Biomedical applications Uhrich KE, Gupta A, Thomas TT, Laurencin C, of polyurethanes: a review of past promises, present reali- Langer R (1995) Synthesis and characterization of de- ties, and a vibrant future. J Biomater Appl 14: 67-90. gradable polyanhydrides. Macromolecule 28: 2148-2193. Uhrich KE, Thomas TT, Laurencin CT, Langer R Discussion with Reviewer: (1997) In vitro degradation characteristics of poly(anhydride-imide) containing trimellitylimidoglycine. N. Gadegaard: Will the catalysts used for the polymeri- J Appl Polymer Sci 63: 1401-1411. sation cause problems in vivo? VanSliedregt A, vanBlitterswijk CA, Hesseling SC, Authors: The most common catalysts used in preparing Grote JJ, deGroot K (1990) The effect of the molecular polylactides and glycolides are organotin compounds and weight of polylactic acid on in-vio biocompatibility. Adv an example is stannous octoate. This catalyst is also used Biomaterials 9: 207-212. widely for making polyurethanes. There are some reports VanSliedregt A, Radder AM, deGroot K, Van on the toxicity of stannous octoate related to breast im- Blitterswijk CA (1992) In-vitro biocompatibility testing plant safety evaluations (Bondurant et al., 2000). One study of polylactides Part I: Proliferation of different cell types. reports no toxic or carcinogenic effects in a 22 month in- J Mater Sci: Mater Med 3: 365-370. vivo study in rats. Polyurethanes polymerized using stan- Verheyen CCPM, deWijn JR, VanBlitterswijk CA, nous octoate have also been used in other medical im- Rozing PM, deGroot K (1993) Examination of efferent plants. Examples includes cardiac pace makers using lymph nodes after 2 years of transcortical implantation of Pellethane as lead insulation. We have also come across poly(L-lactide) containing plugs: a case report. J Biomed references which state that stannous octoate is accepted Mater Res 27: 1115-1118. by FDA as a food additive (Gilding and Reed, 1979, text Vert M, Christel P, Chabot F, Leray J. Hastings GW, reference; Kim et al., 1992) Ducheyne P, eds. (1984) Macromolecular Biomaterials. In most cases the polymerisation catalysts or initiator CRC Press, Boca Raton. pp 119-142. residues are extremely difficult to remove from the final William DF, Mort E (1977) -accelerated hy- polymer, and accordingly they are left to stay with the poly- drolysis of polyglycolic acid. J Bioeng 1: 231-238. mer. Wong WH, Mooney DJ (1997) Synthesis and proper- Camphorquinone type initiators used in in-situ curable ties of biodegradable polymers used as synthetic matrices polymer compositions have also been used in curing den- for tissue engineering. In: Synthetic Biodegradable Poly- tal fillings. Unlike catalysts, radical initiators are incorpo- mer Scaffolds. Atala A, Mooney D, eds. Burkhäuser, Bos- rated into the polymer structure with some residues remain- ton. pp 51-84. ing as low molecular weight compounds in the final poly- Yaszemski MJ, Mikos Ag, Payne RG, Hayes WC mer. (1994) Biodegradable polymer composites for temporary replacement of trabecular bone: The effect of polymer Additional References molecular weight on composite strength and modulus. Mat Res Symp Proc 331: 251-255. Bondurant S, Ernster V, Herdman R (2000) Silicone Yazemski MJ, Payne RG, Hayes WC, Langer R, Mikos Toxicology. In: Safety of Silicon Breast Implants. AG (1996) Evolution of bone transplantation: molecular, Bondurant S, Ernster V, Herdman R, eds. The National cellular and tissue strategies to engineer human bone. Academic Press, Washington, DC. pp 80-113 Biomaterials 17: 175-185. Kim SY, Han YK, Kim YK, Hong SI (1992) Zang JY, Beckman EJ, Piesco NP, Agrawal S (2000) Multifunctional initiation of lactide polymerization by A new peptide-based urethane polymer: synthesis, bio- stannous octoate/pentaerythritol. Makromol Chem 193: degradation, and potential to support cell growth in-vitro. 1623-1631.

16